Jan Philip Wahle

Research Scientist at the University of Göttingen

I have always wondered what makes us humans intelligent. Coming from first principles, I’ve asked myself: “What makes us special compared to anything else on earth?” Basic materials such as soil, sand, or water can move from one place to another but not in organized ways. Plants can grow in organized ways and even perform complex chemical processes to convert the sun’s energy with carbon and water into storable carbohydrates for survival. That’s pretty impressive. But many animals can do even more than that; they can move in organized ways (robotics), using their muscles to get to places where they can receive even more energy, such as vegetables, fruits, or even other animals. They can even see when fruits and vegetables are ripe (vision) and remember which trees are typically the richest.

A key distinction between humans and animals is not that their robotics is much more advanced or that their vision is superior. A wedge-tailed eagle can see prey from as far as 2km, although its eyes are the same size as those of humans. A key aspect that makes us intelligent is the ability to communicate. Many animals also have complex ways of communication, either through sound or other forms, but their ability is primitive compared to humans. Human’s natural language opens up the dimensions of outer worlds (communicating many complex interactions with others) and inner worlds (reflection and thought). It also allows us to travel back and forth in time and create hypothetical scenarios, a key ability many animals do not possess. That’s why if a zebra got eaten by a lion near the water five days ago, the probability that another one will be eaten there soon is high – the other zebras simply cannot warn it of that previous event.

Because I believe language is key to human intelligence, I have dedicated my professional life to understanding it and making computers learn about it to build intelligent systems. Below is a summary of my path so far in that endeavor.

I am currently doing research at the University of Göttingen in Germany with Terry Ruas and Bela Gipp. I have previously worked at the National Research Council in Canada, where I had a great time working with Saif M. Mohammad. My PhD work looks at how natural language processing research is evolving, what we have succeeded at understanding, and what still remains elusive. I also develop technical contributions to understand whether different texts have the same meaning, which is helpful for many applications ranging from language learning to plagiarism detection.

Specifically, I have looked at questions such as:

  1. Which topics is our field working on? Are we working on diverse problems and solutions or concentrating on specific niches?
  2. How are we influencing the broader exchange of ideas? For example, how much are we engaging with other fields, and how much are we drawing from ideas of the past? 
  3. Who are the main actors in our field, and what interests do they have? Who provides funding to our field?
  4. What do we need to teach machines to understand similarities of texts in the same way that humans do?
  5. How do the above change with time, with new policies, or shifts in the focus and interaction of a field?

My research work has numerous applications in NLP tasks like text summarization, media bias, and misinformation detection. In the area of plagiarism detection, I have focused on developing techniques to identify automatically generated content from models like ChatGPT. I am also interested in providing the community with tools to increase the transparency of this new type of partially generated content.

I have published a fair share of research papers in top-tier NLP and AI venues and wrote some blog posts on these topics. I have also given several talks at conferences, universities, and funding agencies. You can follow me on Twitter and LinkedIn to stay up-to-date with my research or check this site every now and then. Who knows? Maybe you will find a gem for yourself that will bring new ideas to your own research.

When I am not working on research, I enjoy outdoor activities like hiking, skiing, and flying FPV drones, as well as playing chess and table tennis and, in general, spending time with the people I love.